Discuss the issue of naming applied to shared memory regions.

What will be an ideal response?

Memory regions may be given textual names. Two or more processes may then share a region by supplying its name and requesting it to be mapped into their address spaces. Mapped files are examples of this idea, in which the contents of the shared region are maintained on persistent store.

The second naming issue is that of the addresses used to access a shared region. In principle, a shared region may be mapped as different address ranges in different processes. However, it is more convenient if the same address ranges are used, for the region may then contain pointers into itself, without the need for translation. Unfortunately, in the absence of global address space management, some addresses belonging to a particular region might already be occupied in some processes.